Arcium, MagicBlock, Umbra, Darklake... An overview of the privacy DeFi and AI ecosystem being built on Solana.

Written by: Castle Labs (@castle_labs)

Compiled by: AididiaoJP, Foresight News

Solana's Privacy Ecosystem is Still in Its Early Stages

We spoke with Helius CEO @mert to ask about his views on Solana's privacy. In his words, Solana is "a bit behind" in terms of privacy.

So, what should a mature privacy ecosystem look like?

Necessary elements include:

• Formal verification
• No governance
• Immutability
• Open-source code

Solana's unique architecture also brings different privacy priorities compared to EVM chains. Mert specifically mentioned ZK compression: "We can achieve scalability and composability of privacy protocols on Solana without persistent Rollups, at least not when based rollups are available."

In Solana's specific context, the two most relevant verticals for privacy development are Neobanks and Private DeFi. However, in terms of tools and user experience, Solana still has a considerable gap before reaching a fully functional, composable privacy ecosystem.

We also asked Mert for his views on the privacy tech stack. As the report emphasizes, privacy should not be seen as a single technology, but rather as an "ultimate privacy stack" where all primitives ultimately work together. For Mert, the endgame will be a combination of FHE (Fully Homomorphic Encryption) and ZK (Zero-Knowledge Proofs). TEE and MPC are practical in certain use cases, but they do not provide sufficient guarantees in adversarial systems.

Finally, we asked him about his thoughts on Helius Privacy.

Helius Privacy will be developed as a ZK-based UTXO privacy layer on Solana. It will utilize "Zones", allowing individual companies to choose their own trade-offs.

It will also provide a public Zone for all regular users, offering complete anonymity in an immutable and formally verified manner. More details will be announced soon.

Based on the above context, this article focuses on how Solana's privacy ecosystem addresses various privacy challenges.

Private Compute

Currently, there are mainly two providers in this field: @Arcium and @magicblock.

Both are addressing similar problems: privacy computing.

Arcium handles arbitrary data through MPC (Multi-Party Computation). It splits data and distributes it across independent node clusters, which collaboratively compute results without seeing individual inputs. At the same time, Arcium operates as an independent computing network, with final settlement on Solana, where tasks are ordered, network security is maintained, and fees paid.

All this privacy computation happens in Multi-Party eXecution Environments (MXEs)—customizable and parallelized execution environments.

In addition to solving computing issues, Arcium's products serve the broader privacy ecosystem of Solana. They are building a Confidential SPL (C-SPL) token standard that enables confidential tokens, transfers, and transactions on Solana.

We inquired about the source of demand from the Arcium team. Unsurprisingly, the main drivers are payments and encrypted data analysis, with increasing institutional demand, especially in healthcare—enabling training models on encrypted datasets. C-SPL also enables seamless private transfers, further attracting institutional interest.

In terms of data, since the Alpha mainnet launch in early February 2026, Arcium has processed over 900,000 computations and more than 3.5 million transactions, with most of the growth occurring in early May.

Currently, most of the demand comes from early applications like Umbra. In the coming weeks, applications such as ZINC and Crafts will also launch, and we expect demand to increase further.

ZINC is working on cryptographic proof-of-work mining, while Crafts utilizes Arcium for sealed-bid auction fundraising, allowing startups to tokenize a portion of equity in a fair price discovery manner.

"Some exciting new things being built on Arcium include: capital formation using sealed bidding, opportunity markets, cryptographic settlement in prediction markets, and other financial privacy applications."

Many of these use cases are creating new or improved markets where users may not even realize privacy is a factor.

Magic Block addresses privacy computing issues through TEE (Trusted Execution Environment), while Arcium relies on the cryptographic assurance of MPC. Their product operates by having Intel TDX create a hardware-verified black box—Private Ephemeral Rollup (PER), where transactions are aggregated and processed before submitting back to Solana.

MagicBlock helps developers retain these attributes throughout the stack, including: confidentiality (protected state), scalability (high throughput), composability (still interoperate with other Solana programs), and compliance (access control layer).

Although they take different paths to solve their respective problems, both can generate deployable private order books, dark pools, and private DeFi tracks with minimal code changes. This is already evident in the ecosystem above Arcium, covering various sectors like DeFi, prediction markets, and Neobanks.

Private Transfers and Balances

As privacy computing infrastructure like Arcium and MagicBlock matures, use cases based on this infrastructure are also beginning to grow, including private transfers.

@UmbraPrivacy is the first to build on Arcium's MPC infrastructure. Umbra introduces Encrypted Token Accounts (ETAs), which are direct counterparts to Solana's standard Associated Token Accounts, but balances are stored in encrypted form, providing:

• Amount privacy: Transaction amounts encrypted using Rescue cryptography
• Balance privacy: Balances stored in ciphertext
• Association privacy: Full severance of on-chain association between sender and receiver through a shielded pool + ZK proof

Additionally, Umbra provides compliance features that allow users to grant selective view permissions to auditors and compliance systems without exposing their complete transaction history. This is critical for institutional workflows and users wanting to prove fund holdings without revealing detailed transaction histories.

In the privacy wallet space, @theprivacycash and Hush are two others.

Privacy Cash employs a Tornado-style shielded pool to handle SOL: users deposit SOL to generate commitments and add them to a Merkle tree, then use ZK proofs to withdraw to any receiving party, completely severing on-chain associations between deposit and withdrawal addresses.

Hush is inspired by Zcash but adds DeFi utility. Users deposit SOL into Hush's shielded pool and it automatically converts to jitoSOL, passively earning staking rewards and MEV income while remaining private. Within the pool, users can transfer to other Hush participants, receive funds, and trade multiple times without touching Solana's public ledger. Upon exiting the pool, withdrawals are unlinked from original deposits through mixing. The intra-pool transfer fee is 0.01 jitoSOL, and a 50 bps fee is charged for withdrawals. Hush also integrates with Jupiter for private swaps and geoblocks sanctioned regions, maintaining a favorable compliance image among institutional users.

No Onchain Trail

Private transfers solve issues for companies and institutions regarding on-chain payments for employee salaries or private transactions. But to go further, we need to embed privacy into everyday on-chain behavior, especially transactions.

Every order on public AMMs is a signal that frontrunners, followers, and MEV bots can read and exploit. There are already some protocols on Solana addressing this issue.

@encifherio is a privacy-first DeFi interface that routes trades through Jupiter while keeping transaction details private. The team states: "Execution quality remains unchanged as we do not add custom routes, using the same routes and liquidity provided by Jupiter."

It wraps the tokens the user wants to exchange and uses ElGamal encryption for the exchange details. On-chain records only show state changes of the wrapped asset type, sufficient for Jupiter to know they are routing the correct tokens. Transaction counts, parties, participants, and even whether a transaction was executed are handled within a TEE environment (AWS Nitro Enclaves), never publicly broadcasted. This enables large-scale private exchanges.

@VanishTrade approaches from a different angle. They are building privacy trading infrastructure that uses shielded trade routing to protect strategies without leaving an on-chain trail. Unlike encifherio's token wrapping, Vanish routes trades through shielded liquidity. Additionally, Vanish has launched the Vanish Integrity Framework (VIF), supported by Elliptic and Range, embedding safeguards to prevent routing any illegal trades.

Darklake is another competitor in this category, building ZK-native liquidity infrastructure and dark pools. Its zk-AMM is called "Blind Spot Slippage Pool", which adds a submission layer on AMMs to hide slippage data before execution. Searchers cannot read order intentions before the trade lands, but can verify results afterward. This delay asymmetry prevents sandwich attacks while retaining verifiability. They have extended this model to private perpetual contracts (zk-Perps) using the Arcium computation layer, along with a reasoning framework called Zyga that abstracts proof complexities, providing builders with a basis for secure logic and coordination. Recently, they also expanded into an infrastructure protocol that allows applications and users to privately verify, connect, and compute using a "Proof as Intelligence" model.

Private Prediction Markets

Private prediction markets are a more advanced privacy application as users' strategies can easily be copied, thus losing advantages. To address this, protocols are building dedicated dark pools for prediction markets using Arcium infrastructure.

@meleemarkets is building prediction markets that support private order flow. They encrypt the order book through Arcium's MPC infrastructure. Participants can place orders without revealing their direction on the public market until settlement.

The Private AI

As AI agents run more frequently on-chain, will every query and every bit of PII they consume be permanently public?

Loyal answers this with its decentralized, censorship-resistant smart protocol. The protocol is built using Magic Block's ephemeral rollup execution and Arcium's encrypted computing. They are building on-chain AI that protects user data: conversations, queries, preferences, and activities, all stored in encrypted form on Solana with strict access rules. Users own and can export their encrypted conversation history and can self-host the front end without losing any data. Furthermore, Loyal also supports private transactions and fund management, allowing depositors to earn yields in private status.